Verification of miRNA Expression using Nuclease Protection and Targeted Next-Generation Sequencing (NGS)

Monica Reinholz, Qian Liu, Mark Schwartz, John Luecke, Chris Roberts, Bonnie LaFleurHTG Molecular Diagnostics, Inc. | Tucson, AZ | Presented at ESHG 2016

HTG Molecular Diagnostics, Inc. | 3430 E. Global Loop | Tucson, AZ 85706 | (877) 289-2615 | htgmolecular.com

HTG EdgeSeq System Workflow

Sample Range Study§ 5 sample types: plasma, serum, FFPE, brain RNA and cell lines§ Biological samples per sample type with an 8-point, 2-fold serial dilution for each§ 144 total wells randomized across two (2) sample plates

Reproducibility Study§ 3 sample types: plasma, FFPE & brain RNA§ 1 biological sample per sample type § 8 technical replicates per sample plate§ 24 total wells per plate randomized by quadrant (first three columns

of a 96 well plate)

Abstract

Sample Input Dynamic Range

Reproducibility

ConclusionsThe HTG EdgeSeq miRNA Whole Transcriptome Assay allows for highlyreproducible expression with low sample input volumes for both plasma andFFPE samples.

§ Does not require RNA extraction from samples tested

§ Detects expression of >2000 miRNAs in different sample types tested

§ Is linear over a wide range of sample inputs

§ Has average Pearson correlation between replicate pairs of ≧ 0.93

§ Is amenable to small clinical specimens—requires very little sample input (~1-2 mm2 FFPE tissue; 12.5 µL blood)

Work supported by NIH grants R44HG005949 and R43HG005949

Equivalent to size of 1.5 mm-diameter TMA core

For Research Use Only. Not for use in diagnostic procedures.

HTG EdgeSeq, HTG Edge and qNPA are trademarks of HTG Molecular Diagnostics, Inc.Any other trademarks or trade names used herein are the intellectual property of theirrespective owners.

Verification Study/Samples

FFPE

% probes with expression = 0 % probes with expression ≤ 100 % probes with expression ≤ 1000FFPE Plasma FFPE Plasma FFPE Plasma

Proc1 26.2 (14.8,36.9) 12.1 (7.8,20) 89.9 (88.2,91.6) 87.8 (87.3,90.2) 96.9 (96,97.7) 96.8 (96.6,97.4)Proc2 19.7 (12.7,24.7) 18.2 (8.3,22.7) 86.1 (78.9,88.8) 87.2 (84.3,88.4) 95 (92.5,96.2) 96 (95.8,96.1)Proc3 16.5 (12.4,27.1) 16.7 (8.3,22.4) 83.4 (73.1,89.4) 87.5 (86.5,88.4) 94 (90.8,96.8) 95.9 (95.6,96.2)Day1 16.5 (12.4,27.1) 16.7 (8.3,22.4) 83.4 (73.1,89.4) 87.5 (86.5,88.4) 94 (90.8,96.8) 95.9 (95.6,96.2)Day2 17.2 (11.8,19.3) 15.2 (7.7,63.5) 85 (74.2,87.1) 89.3 (88.7,98.3) 94.8 (91.3,95.8) 96.7 (96.4,99.7)Day3 15.3 (9.5,17.3) 8.8 (7.5,16.7) 84.1 (82,84.7) 87.5 (86.6,87.9) 94.3 (93.6,94.6) 95.9 (95.7,96.2)

Cell Line

Detection of miRNA Probes Across Processors and Days

Percent probes with p < 0.05 between two cell lines: U251 (glioblastoma) versus H520 (squamous cell carcinoma of the lung). Sample size 3 replicates of each cell line.

Differential expression was demonstrated down to 312.5 cells/well.

Sample Type Sample InputFFPE 1.5 – 12.5 mm2 of a 4-5 µm sectionPlasma 12.5 µLSerum 12.5 µLCell Lines 1,250 – 5,000 cellsPAXgene 32 µLPurified RNA 6.25 – 25 ng

Recommended Sample Input

Sensitivity

Introduction: miRNAs are short, ~22nt RNA sequences that modulate gene transcription and downstream cell behavior. The HTG EdgeSeq miRNA Whole Transcriptome Assay (WTA) enables users to measure the expression of 2,083 human miRNA transcripts using next-generation sequencing. The extraction-free, sample preparation HTG EdgeSeq chemistry is compatible with formalin-fixed, paraffin embedded tissue (FFPE), plasma, serum, whole blood in PAXgene, cell lines, and purified RNA.

Materials and Methods: Two studies were conducted to characterize the performance of the HTG EdgeSeq WTA assay in FFPE and plasma sample types. The sample input studies evaluated a range of sample input amounts for both plasma and FFPE sample types. Optimal input range was established based on sample quality (using process controls) and performance (technical correlation and detection of low expression). Reproducibility studies across technical replicates were also performed for both sample types.

Results: Read depth for FFPE (defined as the total aligned counts at the sample level) ranged from 524K at the lowest sample input (1.56 mm2/well) to 5 million at the highest sample input (12.5 mm2/well) with no loss of sample quality or expression sensitivity. Thus, for FFPE samples, any input volume over the entire range is acceptable. The recommended sample input for plasma was established as 12.5 µL/well; this volume resulted in a minimum of 642K reads. Pairwise, Pearson correlation coefficients on technical replicates ranged from 0.93 to 0.98 for FFPE samples and from 0.82 to 0.98 for plasma.

156.25 cells/well

312.5cells/well

625cells/well

1,250cells/well

2,500 cells/well

5,000cells/well

10,000 cells/well

20,000 cells/well

NA 1639/2280 (72%)

1903/2280(83%)

1696/2280(74%)

1723/2280(76%)

1635/2280(72%)

1693/2280(74%)

1997/2280(88%)

Percent probes detected in 8 technical FFPE and plasma replicates run over 3 processors on a single day and one processor over 3 days (40 replicates per sample type). The table below shows the percent probes with zero counts, ≤ 100 counts, and ≤ 1000 counts, along with the range (highest percentage of probes and lowest percentage of probes), average across the averaged across the 8 FFPE and plasma sample replicates.

Percent probes detected are similar between FFPE and plasma sample types. The percentage of probes expressing moderate (≤ 100) and high (≤ 1000) counts is very similar. Lower expression is more variable between both sample types.

Correlation between replicate samples was summarized by the percentage of pairs with Pearson correlation > 0.90 for FFPE and plasma sample types.

Plasma (196 commonly expressed miRNAs) FFPE#

Wells#

Combinations% with Pearson

≥ 0.90#

Wells#

Combinations% with Pearson

≥ 0.90Proc1 Day1 4 6 100.0% 7 21 100.00%Proc1 Day2 8 28 100.0% 8 28 100.00%Proc1 Day3 8 28 100.0% 8 28 100.00%Proc2 Day1 8 28 100.0% 7 21 100.00%Proc3 Day1 8 28 100.0% 8 28 100.00%Average 36 118 100.0% 38 126 100.00%

Representative replicate correlation plots for A) FFPE (ovary) and B) pooled plasma samples.

A. B.

A. B.

Demonstration of titration performance in A) correlation between representative replicate samples and B) average log2(CPM) expression by quartile for both ovary FFPE and a sample of H520 cell line (non-squamous cell carcinoma of the lung).